Messages in this thread |  | | | Date | Fri, 4 Sep 1998 20:31:13 -0500 (CDT) | | From | "Matthew G. Marsh" <> | | Subject | Re: Bios32.c and DK440LX |
| |
(Doug: I will cut you out of the CC list next round - I was not sure if you needed to be in this ring - mgm)
On Fri, 4 Sep 1998 tmuller@agora.rdrop.com wrote:
> Hi all, > > I've looked at what Matthew had done and something along those lines seems > reasonable to fix the DK440LX's IRQ 9 problems and allow other boards to not > have to have IRQ 9 marked unusable. > > But, I've also noticed that even with the patches that Matthew has, I still am > seeing my Ethernet card on IRQ 9. Bios32.c tells me the following:
I appreciate the attribute 8-) but Mingo actually wrote the patch for Heikki's DK. She posted it to myself and I forwarded it on.
As far as the Ethernet thing - It happened between 117 -> 118. I suspect that one of the IRQ simplifications ends up stripping out the high bits (IE: XT style IRQ vs. AT) because the ethernet always was on IRQ 19 previously. And of course Mingo's patch is still in io_apic.c so on DK440LX boards we end up sharing IRQ 9 between ethernet built-in and the PIIX4 chip. This allows the DK to boot and load drivers but when you go to open the ethernet card you get:
kernel: eth0: speedo_open() irq 9. kernel: eth0: Done speedo_open(), status 00000090. kernel: eth0: Transmit timed out: status 6050 command 0000. kernel: eth0: Tx timeout fill index 15 scavenge index 0. kernel: Tx queue 0001a000 0000a000 0003a000 0003a000 0003a000 0003a000 0003a000 0003a000 000ca000 000ca000 0003a000 000ca000 000ca000 0003a000 400ca000 00000000. kernel: Rx ring 0000a002 0000a002 0000a020 0000a020 0000a020 0000a020 0000a020 0000a022 0000a002 00000001 00000001 00000001 00000001 00000001 00000001 c0000002. **** REPEAT FOLLOWING UNTIL REBOOT **** kernel: eth0: Trying to restart the transmitter... kernel: eth0: Transmit timed out: status 6050 command 0000. kernel: eth0: Tx timeout fill index 15 scavenge index 0. kernel: Tx queue 0001a000 0000a000 0003a000 0003a000 0003a000 0003a000 0003a000 0003a000 000ca000 000ca000 0003a000 000ca000 000ca000 0003a000 400ca000 00000000. kernel: Rx ring 0000a002 0000a002 0000a020 0000a020 0000a020 0000a020 0000a020 0000a022 0000a002 0000a002 00000001 00000001 00000001 00000001 00000001 c0000002. **** REBOOT (C-A-D) SYSTEM REBOOT WORKS **** [snip of very good arguments]
I would agree here on the mapping issue. I have not had time to do more that scan over the patch changes from 117 -> 118.
Mingo - If you would like the complete boot log for 120pre3 vs. 117 with all the APIC details for my machine let me know. I was going to extract some of the points of interest (similar to what Troy had posted about a week ago WRT the PCI mapping lines changing) and post them but I am heading off to Chicago on Tuesday (work...its a living) and won't have time to dig in. Let me know and I can attach them to you. I will be back the following week so if you would like me to test/sample/etc I can do so then.
> > -- > Thanks, > Troy
Thanks to all!
-------------------------------------------------- Matthew G. Marsh, President Paktronix Systems, LLC 1506 North 59th Street Omaha NE 68104 Phone: (402) 932-7250 Email: mgm@paktronix.com WWW: http://www.paktronix.com --------------------------------------------------
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html
|  |